The Long Term Evolution (LTE) project is the 3 G evolution, and it improves and strengthens the 3 G air access technology, uses the Orthogonal Frequency Division Multiplexing (OFDM) and the Multiple-Input Multiple-Out-put (MIMO) as a sole criterion of its wireless network evolution. Peak rates of 326 Mbit/s in downlink and 86 Mbits/s in uplink can be provided in the 20 MHz spectrum bandwidth, and the LTE improves the performance of users at cell edges, improves the cell capacity and reduces the system delay.
The LTE system uses the inter-cell interference suppression technique to improve the data rate at cell edges. Currently, the inter-cell interference technologies mainly comprise Inter Cell Interference Coordination (ICIC), Inter Cell Interference Randomization (ICIR), Inter Cell Interference Cancellation (ICICc) and slow power control, and so on. In the ICIC scheme, the network layer needs to understand the Interference Over Thermal (IOT) in frequencies, and determines whether there is an interference and how the interference strength is by a determination threshold, and then avoids the occurrence of interferences by coordination between base stations as much as possible, wherein the IOT is defined as follows:IOT=NI/N  (1)
where, N is noise of a machine, and NI is a noise interference.
The LTE protocol needs to measure the noise interference power of each resource block and the noise of the machine.
An algorithm for measuring the noise of the machine in the related art uses the following formula for calculation:PThermal=KTB+noise figure  (2)
where, KTB represents the noise, K is the Boltzmann constant, T is a reference temperature (Kelvin), and B is an effective noise broadband of a receiver. In the case of a fixed bandwidth, the value in the algorithm changes with the temperature, the receiver is unable to accurately know the outside temperature, and in the case of a large temperature variation, the algorithm has the problem of measurement inaccuracy.
Another algorithm for measuring the noise of the machine in the related art is to use frequency-domain resources that are not used by users to measure the noise late at night when there are fewer users to access. The algorithm has significant limitations, and needs to determine whether the users are few at this moment artificially.